Designing by Research
By Nelson McKeeby
An Idea Had to Survive Contact with the World
One of the habits that defined Negentropy from the beginning was that an idea did not become part of the setting merely because I liked it. I would imagine something, then go looking for information about how it might work. If the research supported the idea, I kept it. If the idea could work only after some adjustment, I changed it. If the underlying problem was too severe, I discarded it. That process started in 1979 and became one of the main ways the game grew.
I was writing very quickly and constantly, so the supply of ideas was never the limiting factor. The limiting factor was deciding which ideas deserved to survive. Research became a filter. A machine could be interesting, but if it required power that the characters could not generate, that mattered. A weapon could look useful, but if it was too difficult to maintain or supply with ammunition, that mattered. A settlement could seem plausible until I worked out how much food or water it required. A magical effect could be dramatic until it produced a physical result that I did not think matter and energy could actually support. The research did not remove imagination from the game. It gave imagination something solid to push against.
This was especially important because Negentropy was built around people entering a future they did not understand. If every practical difficulty could be solved by inventing another piece of technology, the future would stop being uncertain very quickly. The game needed limits that remained limits after the characters arrived. That made engineering, logistics, food, medicine, power, communication, and transportation part of the story instead of background decoration.
Research Was Part of Writing
I did not separate research from creative work. Reading a Scout pamphlet, an Army manual, a science book, a weapons reference, or a map often produced as much game material as sitting down to write fiction. I might begin with a question raised by an adventure and end with an article explaining a real subject. I might begin with a technical book and discover that one of its examples suggested a machine or a problem that belonged in the setting.
This created a loop that I used repeatedly. A story or game problem produced a question. I researched the question. The research changed the answer and often produced several new questions. Those questions became articles, equipment, places, or later adventures. Because Negentropy used real knowledge as part of play, almost any useful subject could enter the project if it affected what a Troubleshooter might need to understand.
The important point was that research was not there to decorate a decision I had already made. It could overturn the decision. I discarded many ideas because the thing I wanted did not work well enough after I looked at the details. I also kept ideas that became less impressive after research because the limitation made them more useful to the game. The Pipe Car became one of the clearest examples.
The Pipe Car
The Pipe Car began with a simple requirement. Troubleshooters needed a compact vehicle that could fit within the limitations imposed by the Negentropic Cylinders and still move people and equipment after emergence. I wanted something electrically powered because electricity could be generated in several ways and because an electric drive did not require the team to carry a long-term supply of one specific liquid fuel. At the same time, the vehicle had to remain small enough to be believable as Cylinder cargo.
I did not design it by deciding how fast I wanted it to go and then inventing specifications that produced that speed. I went to stores and weighed components. I weighed different kinds of batteries available to me and compared their mass to the energy they could plausibly store. I weighed electric motors. The first Pipe Cars used four motors because that arrangement gave me a practical way to think about distributed drive and available power while keeping the machinery compact.
A science teacher helped me work through the relationship between mass, speed, and power. My calculations were simple by engineering standards, but they forced the design to answer real questions. A heavier vehicle required more energy to accelerate and more power to maintain useful speed. More batteries added energy but also added mass. Bigger motors could provide more power but occupied more space and increased the electrical demand. Every improvement created another cost somewhere else.
This was exactly the kind of problem I wanted Negentropy to preserve. The vehicle did not have to be optimized like a commercial automobile. It had to be understandable enough that a player could see why it performed the way it did and why a different choice would change something else.
X-Tech Was an Allowance, Not an Escape
I allowed myself one fictional adjustment. I called it X-Tech. The idea was that aliens or an advanced company could produce limited quantities of technology that was ahead of what ordinary industry could manufacture. I used it as an excuse to push some capabilities beyond the equipment I could actually buy or find in catalogs, but I tried to keep the improvement within a range that still left the original engineering problem recognizable.
For batteries and motors, I often treated X-Tech as roughly doubling performance or capacity. That was a large advantage, but it was not magic. A battery with twice the energy density was still a battery with finite energy. A motor with improved output still required power and produced heat. A computer could be smaller and more capable than the machines I knew, but it still had to communicate, store information, and draw electricity. X-Tech let me imagine the direction technology might go without allowing every problem to disappear.
This distinction became important later because players naturally wanted the best equipment they could imagine. If a modest improvement was acceptable, why not a much larger one? If a compact battery could be twice as good, why not make it a hundred times better? If an electric car had limited range, why not give it a quantum power source? Those suggestions made sense from the point of view of players who wanted their characters to be more capable. I rejected them because the constraint itself was part of the game.
Why I Refused Quantum Power
The early Pipe Cars were not very powerful. They also consumed electricity rapidly. Even after applying the X-Tech assumption, their performance remained limited by the relationship among battery mass, motor power, vehicle mass, and available space. Players repeatedly suggested that the obvious answer was some kind of advanced power system. Quantum power was one of the ideas proposed. Another recurring suggestion was simply to make the Cylinders larger so that the Project could send heavier vehicles and more energy storage.
I refused both solutions because they solved the wrong problem. The point was not to give Troubleshooters the best possible vehicle. The point was to ask what useful vehicle could be sent under the conditions already established. If the Cylinder had finite volume, then equipment selection had to involve tradeoffs. If electricity had to be generated and stored, then mobility had a cost. If the car used too much power, the team had to decide when driving was worth the expense.
Those decisions generated play. A character who could drive anywhere indefinitely did not need to think about charging, routes, weight, or whether the vehicle should be risked. A character with limited electrical energy had to consider all of those things. The Pipe Car became more interesting because it was not powerful enough to erase the environment.
This was one of the first places where I saw a recurring difference between what players often wanted and what I wanted the setting to preserve. Players tended to see a limitation and ask what technology would remove it. I tended to see the same limitation and ask what decisions it would force.
Cinematic Reality and Concrete Reality
The distinction eventually became part of the postmortems we held after early games. We talked about cinematic reality and concrete reality. Cinematic reality asks whether something works for the scene. If a vehicle needs to make a dramatic escape, it finds enough power. If a gun needs another shot, perhaps there is another round. If a radio needs to reach someone, the signal gets through. The story is organized around the action the scene requires.
Concrete reality asks a different question: what is physically present and what can it actually do? If the battery is depleted, the car does not move because the next scene would be more exciting if it did. If the characters brought the wrong tool, the missing tool remains missing. If the terrain blocks the radio path, the communication problem must be solved another way. The facts of the situation continue to exist when they become inconvenient.
I did not think Negentropy had to exclude cinematic play. Role-playing is storytelling, and dramatic compression is useful. The problem came when cinematic convenience erased the practical questions that defined the game. If the characters were supposed to be Troubleshooters, then equipment failure, limited supply, difficult terrain, and incomplete information were exactly the kinds of things they were supposed to troubleshoot.
Research gave concrete reality a foundation. It meant I could explain why a limitation existed instead of simply asserting that the game master had decided to make something difficult.
The Value of Being Wrong
Research also made it possible for me to be wrong in a productive way. I could begin with an assumption, discover that it did not survive calculation or comparison, and then rebuild the idea. That became normal. I did not regard an abandoned design as wasted work. Finding out that something failed often produced a better understanding of what the setting actually needed.
This happened with technology repeatedly. A proposed machine might require too much energy, so I had to decide whether the machine should become smaller, slower, less capable, or disappear. A weapon might be effective but create ammunition and maintenance problems that made another design more suitable for Project use. A communications system might look impressive until I considered range, power, antenna requirements, or what happened when the network infrastructure was gone.
The same method applied outside engineering. Agriculture depended on season, soil, water, labor, and time. Medicine depended on anatomy and physiology rather than a generic healing rule. Settlements required food and sanitation. Armies required supply. Travel required maps, weather, terrain, and energy. Research turned each subject into a set of connected constraints, and those connections made the world more coherent.
The Near-Possible Technology Rule
Over time I became increasingly comfortable with technology that was advanced but still recognizable. I did not need Project equipment to look primitive, but I wanted the player to understand what problem each device solved and what limitations remained. A better battery was useful because everyone already understood why a battery mattered. A compact digital radio network was useful because the basic functions of radios and computers remained familiar even if the Project version was far beyond the TRS-80 my school had recently acquired.
That is what X-Tech was best at. It could move a known technology forward without severing the connection to the real thing. A Project radio could be smaller, more capable, digitally coordinated, and integrated with computers, while still depending on power, antennas, distance, and signal conditions. A Project computer could be vastly better than contemporary personal computers while still being a computer rather than an oracle. Improved materials could make equipment lighter or stronger without making mass irrelevant.
I preferred this to ray guns, magical phasers, unlimited reactors, and other devices whose main function was to remove constraints. Those technologies could be enjoyable in another kind of science fiction, but they made it harder to ask the practical questions that interested me.
Research and the Project Arsenal
The same method shaped the original Project weapons. I did not begin by creating a list of futuristic guns. I started from equipment that I could research and from people who had actually used it. The former French naval serviceman I interviewed after his work with Jacques Cousteau's Calypso influenced the first Logi because he treated equipment as something that had to remain useful after long service. His rifles, pistol, Uzi, diving equipment, outboards, and other tools were not props. They were maintained because people depended on them.
I then used references such as Joseph W. Smith's Small Arms of the World to verify and develop the weapons I had recorded in my notes. The MAS rifles, Browning/FN 1935, revolvers, and ordinary 12-gauge police shotgun were kept because they could be understood as durable tools rather than fashionable weapons of the moment. Research sometimes changed the exact model, as when the rifle I had recorded as a MAS 1944 led me toward the MAS 1949/56, but the principle remained the same.
This was another place where player preference tested the method. Some players later wanted the AK-47 because of its reputation as the greatest rifle in the world. I was less interested in selecting the weapon with the strongest reputation than in asking what a civilian Project would deliberately standardize. The M16, AKM, FAL, and G3 were all strongly tied to the major Cold War powers. The French rifles and Uzi felt less politically defining to me, especially after de Gaulle's withdrawal from NATO's integrated military command. Whether every historical judgment I made as a young designer was perfect is less important than the method: the equipment had to make sense inside the institution that selected it.
Research Made Equipment Part of Character
This also changed the meaning of equipment in play. A piece of gear was not simply a bonus printed on a character sheet. It was something with capacity, mass, maintenance needs, operating requirements, and a reason for being there. If a Troubleshooter knew how to use it, that training mattered. If the Project expected the item to remain useful decades or centuries after it was packed, reliability mattered. If the Cylinder could not be reopened to add a newer model later, the original choice mattered even more.
That permanence was one of the most useful consequences of the Cylinder premise. The Project had to choose before it knew exactly what conditions the team would face. The designers could not chase the gun of the month, the newest electronics, or whatever machine happened to be fashionable at the moment of play. They had to select things that were broadly useful and teach people to use them before departure.
Research therefore connected equipment to institutional planning. I had to think about what a Project quartermaster, engineer, doctor, or communications specialist would actually choose when every kilogram occupied finite space and every replacement might be unavailable after emergence. The setting became stronger when the equipment list looked like a set of decisions made by an organization rather than a catalog assembled for players.
Research as a Defense Against Fashion
One of the risks in a long-running science-fiction game is that current technology constantly changes. Something that appears advanced one year can look ordinary a decade later. Something fashionable can disappear. A game that continually replaces its equipment with whatever is newest can lose any sense that the fictional organization made decisions of its own.
The research method helped resist that. I could ask whether a new technology actually solved a Project problem better enough to justify changing doctrine, training, supply, and maintenance. Sometimes the answer was yes. Computers and digital communications clearly belonged in the direction technology was moving, so X-Tech pushed them far beyond the systems I could see at school or read about. In other areas, a mature mechanical solution remained good enough. A pump shotgun did not need to become exotic simply because science fiction allowed it.
This gave the Project a technological personality. Some equipment was remarkably advanced. Other equipment was conservative. The difference depended on whether improvement changed the mission enough to justify it. That mixture became more believable to me than a future in which every object had been replaced by a futuristic equivalent.
The Research Had to Be Understandable
I also wanted the research to remain usable by players. There was no advantage in building a technically elaborate system that nobody at the table could understand. The point was to make the physical situation clearer, not to bury it beneath equations. When I used calculations, manuals, or technical references, I tried to reduce them to the principle that controlled the decision.
For the Pipe Car, the important relationship was easy to understand even without detailed engineering mathematics. Batteries stored a finite amount of energy. Motors required power. More vehicle mass required more energy to move. More batteries increased both stored energy and mass. Cylinder volume limited how large the whole system could become. Once those facts were established, players could reason about the vehicle.
That level of explanation became a model for later sourcebooks. The text should teach enough of the real mechanism for players and game masters to make informed decisions. It did not need to reproduce a professional engineering manual. It needed to explain why one choice produced one consequence and another choice produced something else.
Research and Magic
The same habit reached even into the magical realism of the setting. I allowed magic to manipulate energy and alter the arrangement of matter, but I did not want it to create final states that were physically impossible. That meant magic was another extraordinary mechanism operating against ordinary consequences.
If a magical working heated something, the heat still existed. If matter was rearranged, the resulting material still needed a physically possible structure. If a healing effect changed tissue, the body still had to contain viable tissue in a configuration that could function. The mechanism could be impossible by contemporary science, but the result could not simply ignore matter, energy, anatomy, or chemistry.
This rule let magic coexist with research instead of replacing it. A character who understood medicine, engineering, or materials science could still reason about a magical result because the world after the effect remained a physical world. That was more interesting to me than using magic as permission to stop asking how anything worked.
The World Became an Argument I Could Test
As Negentropy grew, I came to think of many setting decisions as arguments. A town existed because there was a reason people could feed themselves there. A trade route existed because something worth moving could be transported over it. A machine had the capabilities I gave it because its power and materials could plausibly support them. A military organization carried certain equipment because it could manufacture, supply, and maintain it.
Research did not prove that the fictional world was real. It gave me ways to test whether one part contradicted another. If a riverboat supposedly carried a certain load, I could compare displacement and available power. If a region supported a population, I could look at agriculture, rainfall, transport, and storage. If an army had thousands of soldiers, I could ask what they ate and how ammunition reached them. Every answer created another constraint, and the network of constraints made the setting more resistant to arbitrary change.
That resistance is valuable. A world that changes instantly whenever a writer wants a convenient answer never develops much internal structure. A world in which earlier decisions continue to have consequences becomes easier to believe and, for a game, easier to explore.
Why I Still Design This Way
The tools available for research have changed enormously since 1979, but the basic method has not. I still begin with questions. I still look for the real mechanism before deciding how much fiction is necessary. I still prefer to improve or extrapolate an existing technology when that will solve the problem, and I still discard ideas when the research shows that keeping them would damage more of the setting than they add.
The most important part of the method is not realism for its own sake. It is consequence. If something has mass, power requirements, cost, maintenance, training demands, biological effects, or logistical needs, those properties create decisions. Decisions create play. Removing every limitation usually removes the decision with it.
The Pipe Car survived because it was useful without being perfect. It could move people and equipment, but not indefinitely. It could be improved by X-Tech, but not freed from power consumption. Its small size was a consequence of the Cylinder, and the Cylinder remained small because finite space was one of the premises that made Project planning meaningful. Players repeatedly proposed better answers, and many of their suggestions improved the game, but the proposals that erased the underlying constraint were the ones I was least willing to accept.
That is what designing by research has meant throughout Negentropy. I am willing to invent the impossible mechanism when the setting needs one. After that, I want the world to behave as though consequences are real. I imagine the thing, find out what it would require, and then let the answer change the design.